Humans can make abstract choices independent of motor actions. However, little is known about the functional role and neural representation of abstract choices. Here, we show that in the human brain choices are represented in an abstract, motor-independent manner, even when they are directly linked to an action. To disentangle sensory, choice, and motor aspects of decision-making, we measured MEG signals while human participants made choices with known and unknown motor response mapping. Using multivariate decoding, we found stimulus, choice and response information with distinct cortical distributions. Choice representations were invariant to whether or not the response mapping was known during stimulus presentation. Furthermore, neuronal choice representations predicted decision confidence and occupied distinct representational spaces from both stimulus and motor signals. Our results uncover abstract neuronal choice signals that generalize to embodied decisions. This suggests a general role of an abstract stage in human decision-making.
bioRxiv Subject Collection: Neuroscience